Was SA’s 2024 third-quarter GDP contraction caused by an overestimation of agriculture’s decline?
By Ed Stoddard
The shock 0.3% quarter-on-quarter contraction in South Africa’s economy in the third quarter (Q3) of last year stemmed from a more than 28% quarter-on-quarter slide in agricultural production.
But the Bureau for Food and Agricultural Policy (BFAP) has questioned the scale of that decline in Statistics South Africa’s (Stats SA’s) data, sowing seeds of doubt about the overall shrinkage in gross domestic product (GDP).
This is material with consequences that extend far beyond the platteland. If South Africa’s gross GDP contracted again in Q4 — that data will be released in early March — the economy will have fallen into a recession. And the dreaded “R-word” will have market-moving implications and sour investment perceptions.
The agricultural slump shaved 0.7 percentage points off the Q3 GDP number. If the sector was removed from the equation, the economy would have posted muted growth of 0.4%. And if agriculture’s contraction was overestimated, the economy may have flatlined or grown by, say, 0.1%.
This may not seem like a massive deal, but the difference between the “C-word” (contraction) and the “G-word” (growth) has huge implications when it comes to GDP and how investors regard an economy.
The issue is partly a matter of timing, which can heavily skew quarter-on-quarter comparisons.
“What happened with the 28% decline was that you had a whole lot of the maize crop that was delivered in June as opposed to July. And then next year it might be the other way around,” Dr Tracy Davids, the BFAP’s executive director of commodity markets and foresight, told an agricultural roundtable hosted by Nedbank on Tuesday.
June falls in the second quarter while July is the first month of the third quarter.
“So, you have this massive swing quarter-on-quarter which is just a timing issue,” said Davids.
Because of such gyrations, the BFAP prefers to measure year-to-date output from the sector compared with the same period in the previous year. And in fairness to Stats SA, it needs to crunch and provide the quarterly, seasonally adjusted numbers.
But even by the year-to-date readings, the BFAP’s calculations are significantly at odds with Stats SA’s.
A gigantic gap
The BFAP conducted a rapid assessment of the data available to Stats SA in the wake of the Q3 assessment at the request of Agri SA and the Agricultural Business Chamber of South Africa (Agbiz).
These organisations found the extent of the decrease — even accounting for last summer’s El Niño-triggered drought — went against the grain of their observations.
The BFAP published a report on its findings in late December when South Africa was deep in holiday mode.
“The analysis indicates that the year-to-date (first three quarters of 2024) decline in real agricultural GDP should be between 5%-6%, as opposed to the current official decline of 15.5%,” said the report.
That is a gigantic gap and for the overall GDP read can mean the difference between the C-word and the G-word on a quarterly basis. It is also a hot potato for the year-to-date calculations to the end of Q3.
“Our adjustments indicate that an upward revision of approximately R10.8-billion is necessary for the real agricultural GDP. If this estimate is incorporated into the South African economy, it suggests that the economy would have grown year-to-date in real terms by 0.7%, rather than 0.4%,” said the report.
It noted that this revision was more in line with forecasts modelled by the South African Reserve Bank and the Bureau for Economic Research.
“While acknowledging the complexity and difficulty in sourcing real-time statistics to calculate the sector’s economic performance in aggregate, the report suggests that reforms are necessary to ensure more robust indicators for the agricultural economy,” said the report.
Areas of concern highlighted by the report included “a significant discrepancy in horticultural income levels”, which the BFAP maintains was underestimated in the official data.
“Additionally, our adjustments to the expenditure side add several billion rands of value,” said the report. This all throws into sharp relief the difficulties in measuring agricultural output.
“Calculating South Africa’s agricultural GDP is a difficult undertaking due to the complex nature of the various agricultural industries, their linkages to the rest of the economy and the seasonal nature of farm production,” said the BFAP report.
More volatile
“This means that agricultural GDP, unlike many other sectors of the economy, is much more volatile and exposed to factors which influence the various components of the calculation of GDP.”
Stats SA can only work with the data provided to it. In response to Daily Maverick’s queries, it said that the issue stemmed from its input data, but work was under way on that front.
“We have not changed the methodology, and we are still continuing with the current methodology as it remains valid. The issue is with the input data, which Stats SA can’t improve since it is acquired from the external source [Department of Agriculture],” said Bokang Vumbukani-Lepolesa, the chief director of national accounts at Stats SA.
“I believe the Department of Agriculture is working on improving the data and we are yet to hear what is being done, or at least what the plan is towards improvement of the data. If we receive the revised input data, we will revise the figures for Q3.”
The Department of Agriculture had not responded to our queries by the time of publication.
Stats SA frequently revises its initial estimates and may do so when the Q4 2024 GDP data are released in early March — if it gets revised input data. This is standard practice for statistical agencies worldwide.
And if South Africa’s agricultural output has been underestimated, it is surely good news that the economy is in better nick than the official data currently shows. That is a green shoot of note.